[Paper Review] Detecting and modelling real percolation and phase transitions of information on social media

Jiarong Xie, Fanhui Meng+4 more|arXiv (Cornell University)|Mar 4, 2021

Complex Network Analysis Techniques64 references4 citations

TL;DR

This study analyzes 100 million Weibo and 40 million Twitter users to detect real percolation and phase transitions in information spread on social media. It reveals a lower percolation threshold than predicted by theory, driven by positive feedback in the coevolution of network structure and user activity, leading to extreme influence imbalances and higher-than-expected information diffusion capacity.

ABSTRACT

It is widely believed that information spread on social media is a percolation process, with parallels to phase transitions in theoretical physics. However, evidence for this hypothesis is limited, as phase transitions have not been directly observed in any social media. Here, through analysis of 100 million Weibo and 40 million Twitter users, we identify percolation-like spread, and find that it happens more readily than current theoretical models would predict. The lower percolation threshold can be explained by the existence of positive feedback in the coevolution between network structure and user activity level, such that more active users gain more followers. Moreover, this coevolution induces an extreme imbalance in users' influence. Our findings indicate that the ability of information to spread across social networks is higher than expected, with implications for many information spread problems.

Motivation & Objective

• To investigate whether real percolation and phase transitions occur in information diffusion on social media.
• To identify the mechanisms underlying unexpectedly low percolation thresholds in real-world social networks.
• To model the coevolution between user activity levels and network structure in shaping information spread dynamics.
• To quantify the influence imbalance caused by feedback loops in social media networks.

Proposed method

• Analysis of large-scale social media data from Weibo (100M users) and Twitter (40M users) to track information cascades.
• Application of percolation theory to detect phase transitions in information diffusion across social networks.
• Modeling the coevolution of user activity levels and network structure using dynamic feedback mechanisms.
• Quantitative comparison of observed percolation thresholds with theoretical predictions from classical percolation models.
• Use of statistical physics frameworks to interpret information spread as a phase transition process.
• Identification of positive feedback loops where active users gain more followers, accelerating network growth and information spread.

Experimental results

Research questions

• RQ1 Do real percolation and phase transitions occur in information diffusion on social media platforms like Weibo and Twitter?
• RQ2 Why is the observed percolation threshold lower than predicted by classical percolation theory in social networks?
• RQ3 How does the coevolution between user activity and network structure affect the spread of information?
• RQ4 To what extent do feedback mechanisms amplify the influence of highly active users in social media?
• RQ5 What are the implications of these dynamics for the overall capacity of social networks to propagate information?

Key findings

• The observed percolation threshold in real social media networks is significantly lower than theoretical predictions, indicating easier information spread than expected.
• Positive feedback in the coevolution between user activity and network structure drives the lower percolation threshold.
• Highly active users rapidly accumulate followers, creating extreme influence imbalances that amplify information diffusion.
• The coevolution mechanism leads to a super-linear growth in user influence, accelerating cascade formation.
• Information spreads more readily in real networks than in static or random network models due to dynamic structural feedback.
• The findings suggest that social media networks have a higher inherent capacity for large-scale information propagation than previously modeled.